President Bush’s proposal to eliminate individual income taxes on dividends provoked several hasty and misleading reactions — even from supporters of the plan.
To grapple with this tricky topic it helps to understand there are only a few things a corporation can do with its earnings. Before any listed corporation can even think about paying a dividend, creditors and tax collectors have to be paid. Even before corporate taxes are paid, interest has to first be paid to bondholders and banks. That interest expense (unlike dividends) is as deductible as any other cost. Since the cost of repaying stockholders with dividends is not deductible, however, tax laws have made it cheaper to finance investments with debt rather than by selling shares. Indeed, many companies have been buying back their own shares while piling up debt. Maximizing the ratio of debt to equity has minimized corporate taxes quite effectively but it has also imposed painful bankruptcy costs on the whole economy.
Corporate bonds, unlike dividend-paying stocks, are taxed only once — at the individual level. Each dollar of corporate profits distributed as interest payments to bondholders is taxed once at the 35% corporate rate, leaving 65 cents. A dollar distributed as dividends to stockholders is also reduced to 65 cents, but that 65 cents is then taxed a second time, leaving only 40 cents for top-bracket taxpayers (closer to 30 cents after subtracting state taxes).
Larry Kudlow recently wrote that “shareholders will keep 100 cents on each new dividend dollar, compared with only 65 cents on each dollar of interest paid on corporate bonds.” To non-economists that probably sounded as though the Bush plan treats corporate bondholders more harshly than stockholders, but it actually treats them the same if corporate and individual tax rates are about the same.
Interest is a deductible cost for corporations and therefore taxable income for individuals. Dividends are not deductible for corporations, so they would not be taxable income for individuals under the Bush plan. With no double tax on dividends, profits used to pay interest or dividends will both be taxed at 35% for top-bracket taxpayers.
Either making dividends totally deductible from the corporate tax or making dividends totally tax-free at the individual level would make the tax system neutral between debt and equity. As a political matter, however, it is hard to imagine Congress allowing such a generous new deduction for “big business.” Moreover, as a practical matter, it also turns out to be best for corporations to provide major dividend relief at the individual level.
To instead make both interest and dividends completely deductible at the corporate level, as advocated by people as diverse as columnist George Will and Sen. Jon Corzine, would be far more problematic. With such large deductions, many companies would have insufficient taxable earnings against which to write-off major investments in buildings and machines in a reasonable period of time (in tax jargon, capital-intensive firms could wind up with excessive tax losses to carry forward). The inability to fully write-off capital costs in a timely way would have the opposite effect of accelerated depreciation — that is, it could easily delay or discourage business investment.
The debt-equity distortion is only one way the tax system causes corporations and their shareholders to invest inefficiently. Corporate decisions about whether or not it would be best to reinvest all after-tax profits or pay some of them out as dividends are also affected by the double taxation of capital gains.
After corporations deduct frequently excessive interest expense, what remains is taxable income. And after paying federal and state taxes on taxable income, each dollar shrinks to about 60 cents. At that point, there are only two things corporate managers can do with after-tax earnings: They can either pay a dividend or they can retain and reinvest the earnings (though not always wisely).
Total elimination of the tax on dividends would tilt the tax system against double-taxed retained earnings just as the current system tilts the system against paying dividends. Burton Malkiel of Princeton recently wrote that, “if corporate earnings are retained, they are taxed only once.” But that ignores the reason earnings are retained — to produce capital gains.
Unless corporations invest retained earnings very badly the result is more assets per share. Even putting retained earnings in the bank or using them to buy back shares has that effect. With more assets per share, the value of each share normally goes up, resulting in a capital gain. Such capital gains, like dividends, are subject to double taxation of individual stockholders — albeit at a more tolerable 20% rate.
Perfect neutrality would require eliminating the capital-gains tax on stock, but perfection should never be the enemy of progress. The Bush proposal gets at the bias against retained earnings in a roundabout way by allowing investors to treat retained earnings as “deemed dividends” per share and adding that figure to the cost basis of their shares to later lessen the capital-gains tax. This is a quite clever plan in theory but it may prove somewhat difficult to implement and even more difficult to explain, particularly to Congress. It is always difficult to underestimate the capacity of legislators to understand economics.
The much simpler alternative of just making the individual tax on dividends and capital gains identical — 20% — could wind up as the best compromise between those who want to keep the dividend tax sky high and those who hope to scrap it entirely. A 20% tax on both dividends and capital gains would still leave a slight advantage for debt financing, and it would obviously leave a higher burden on saving and investing. On the other hand, a 20% rate would probably raise more revenue than the current dividend tax because it would result in more dividend-paying stock being held outside of tax-free funds, and greater use of equity would shrink corporate interest deductions.
Alternative compromise proposals are vastly inferior. The worst idea is excluding a few hundred dollars of dividends each year. Such an exemption would reduce revenues substantially with no beneficial effects on marginal decisions of companies to pay more dividends — or of taxable investors to hold more dividend-paying stocks. It could easily induce small investors to limit their dividends to the exempt amount, selling shares if dividends grew too large.
Another second-rate option is to exclude half of dividends from taxation — that is, taxing dividends at assorted tax rates ranging from 5% to 18%. That would lose a lot more revenue than a flat 20% dividend tax while also adding more distortions. With this scheme, for example, a thousand dollars of dividends could be taxed much more lightly than a thousand dollars of capital gains, depending on the taxpayer’s other income. It is never good policy to make the marginal tax on investments depend on annual salary or to make the marginal tax on salaries depend on gross income from dividends or capital gains.
I do not believe a 20% tax on dividends would add even a dollar to budget deficits, even aside from its clearly favorable impact on economic growth. On paper, a 20% tax on dividends would supposedly lose only $7-8 billion per year, much less than a routine estimating error. In reality, the lower tax on dividends would result in larger taxable capital gains, reduced corporate deductions for interest expense, and a larger share of bigger dividend payments being made to taxable investors rather than to tax-free pension funds and foundations.
If we start out asking for zero tax on dividends and end up getting 20%, that is not necessarily a setback if the resulting reduction of budgetary constraints (at least $40 billion a year) is redeployed for other growth-oriented tax reforms. A few excellent additions would be to cut the estate tax to 20% right away, roll the top income-tax rate back to 31% for both individuals and corporations, and scrap the corporate and individual alternative minimum tax. Accomplish even a few of these modest tasks, and federal, state, and local revenues from all sources would soon begin to grow quite vigorously as a consequence of the more vibrant economy and stock market.
— Alan Reynolds is a senior fellow at the Cato Institute and a contributor to National Review.